An Intelligent Optical Lens Assisted Mobile Application System for Early Detection of Melanoma

Abstract

Melanoma is the fatal form of skin cancer that can prevail anywhere in the body due to the abnormal proliferation of melanin pigment. The early detection of Melanoma is difficult due to the absence of anti-natural symptoms it owns. Normal moles may turn into Melanoma lesions due to many factors. A dermatologist diagnoses melanomas using an instrument called Dermatoscope, which acquires the lesion's image and processes it based on the ABCDE algorithm. This study has fabricated an optical lens system that can capture the lesion's appearance and make it free from pixilation. This lens system is a dual-lens system integrated with a mobile application that can predict the possibility of the mole being a lesion and classify it based on its severity using a Deep ConvNet structure. Using an integrated lens system instead of a conventional dermatoscopy is a cost-efficient method and provides clinical significance to identify Melanoma in its earlier stage. Integrated lens apparatus is handy due to its less complicated nature of construction. This setup's efficiency is measured based on its adaptability, output image quality, and accuracy. Detection time and efficiency of the system is assessed using a renowned clinician analyst.